Hello,

kernel test robot noticed "kernel-selftests.seccomp.seccomp_benchmark.fail" on:

commit: bb94019905757de0ddac22e897b7648b55aeef8e ("kselftest/seccomp: Report each expectation we assert as a KTAP test") https://git.kernel.org/cgit/linux/kernel/git/broonie/ci.git kselftest-seccomp-benchmark-ktap

in testcase: kernel-selftests version: kernel-selftests-x86_64-60acb023-1_20230329 with following parameters:

group: group-s

compiler: gcc-12 test machine: 36 threads 1 sockets Intel(R) Core(TM) i9-10980XE CPU @ 3.00GHz (Cascade Lake) with 32G memory

(please refer to attached dmesg/kmsg for entire log/backtrace)

If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags | Reported-by: kernel test robot oliver.sang@intel.com | Closes: https://lore.kernel.org/oe-lkp/202312271441.2e64d159-oliver.sang@intel.com

# timeout set to 120 # selftests: seccomp: seccomp_benchmark # TAP version 13 # 1..7 # # Running on: # # Linux lkp-csl-d01 6.7.0-rc3-00003-gbb9401990575 #1 SMP PREEMPT_DYNAMIC Mon Dec 25 09:40:50 CST 2023 x86_64 GNU/Linux # # Current BPF sysctl settings: # # /proc/sys/net/core/bpf_jit_enable:1 # # /proc/sys/net/core/bpf_jit_harden:0 # # Calibrating sample size for 15 seconds worth of syscalls ... # # Benchmarking 37997130 syscalls... # # 15.768643565 - 1.028667802 = 14739975763 (14.7s) # # getpid native: 387 ns # # 32.305042795 - 15.769439484 = 16535603311 (16.5s) # # getpid RET_ALLOW 1 filter (bitmap): 435 ns # # 48.595567892 - 32.307930690 = 16287637202 (16.3s) # # getpid RET_ALLOW 2 filters (bitmap): 428 ns # # 67.773953317 - 48.596835421 = 19177117896 (19.2s) # # getpid RET_ALLOW 3 filters (full): 504 ns # # 88.068486378 - 67.774726217 = 20293760161 (20.3s) # # getpid RET_ALLOW 4 filters (full): 534 ns # # Estimated total seccomp overhead for 1 bitmapped filter: 48 ns # # Estimated total seccomp overhead for 2 bitmapped filters: 41 ns # # Estimated total seccomp overhead for 3 full filters: 117 ns # # Estimated total seccomp overhead for 4 full filters: 147 ns # # Estimated seccomp entry overhead: 55 ns # # Estimated seccomp per-filter overhead (last 2 diff): 30 ns # # Estimated seccomp per-filter overhead (filters / 4): 23 ns # # Expectations: # # native ≤ 1 bitmap (387 ≤ 435): ✔️ # ok 1 native ≤ 1 bitmap # # native ≤ 1 filter (387 ≤ 504): ✔️ # ok 2 native ≤ 1 filter # # per-filter (last 2 diff) ≈ per-filter (filters / 4) (30 ≈ 23): ❌ # not ok 3 per-filter (last 2 diff) ≈ per-filter (filters / 4) # # 1 bitmapped ≈ 2 bitmapped (48 ≈ 41): ❌ # not ok 4 1 bitmapped ≈ 2 bitmapped # # Skipping constant action bitmap expectations: they appear unsupported. # ok 5 # SKIP entry ≈ 1 bitmapped # ok 6 # SKIP entry ≈ 2 bitmapped # ok 7 # SKIP native + entry + (per filter * 4) ≈ 4 filters total # # Saw unexpected benchmark result. Try running again with more samples? # # Totals: pass:2 fail:2 xfail:0 xpass:0 skip:3 error:0 not ok 2 selftests: seccomp: seccomp_benchmark # exit=1

The kernel config and materials to reproduce are available at: https://download.01.org/0day-ci/archive/20231227/202312271441.2e64d159-olive...